Google co-founder works with AI researchers on powerful model Gemini.
Sergey Brin has been visiting headquarters three to four times a week. He was already attending meetings on AI last year, but according to Wall Street Journal sources, the frequency and intensity is increasing. A turnaround, because after resigning from his executive position at parent company Alphabet, he is said to be taking a more hands-off approach.
During his visits to headquarters, Brin works with researchers on AI model Gemini. Among other things, they discuss technical issues about “loss curves,” or measuring the performance of AI systems over time. He also sets up weekly meetings on new AI research and is involved in personnel matters. For example, he would help attract sought-after researchers.
Brin’s stepping up comes at an interesting time, given the struggles going on at the generative AI level. ChatGPT managed to become enormously popular in a short period of time. Google now has an answer with Bard, but in the meantime, the development of major AI initiatives continues.
With Gemini, Google wants to build a general-purpose AI system that can compete with OpenAI’s GPT-4 model. This model is the basis of a paid ChatGPT version. Top executive Demis Hassabis is said to have since informed Google employees that Gemini will be available this year.
During the recent edition of Google I/O, Google spoke about Gemini. “Gemini was created from the ground up to be multimodal, highly efficient at tool and API integrations and built to enable future innovations, like memory and planning. While still early, we’re already seeing impressive multimodal capabilities not seen in prior models,” Google said.